Loading scripts/benchmark/service_benchmark/navigation/run.sh +1 −1 Original line number Diff line number Diff line Loading @@ -25,6 +25,6 @@ mkdir -p benchmark_results # Run benchmark echo "Running service benchmark" python -m vagen.utils_benchmark.service_benchmark --config "$SCRIPT_DIR/benchmark_config.yaml" python -m vagen.env.verify_service --config "$SCRIPT_DIR/benchmark_config.yaml" echo "Benchmark complete. Results saved to benchmark_results directory." No newline at end of file scripts/benchmark/service_benchmark/svg/dataset_config.yaml +2 −2 Original line number Diff line number Diff line Loading @@ -5,5 +5,5 @@ svg-vision: dino_weight: 1.0 structural_weight: 1.0 dreamsim_weight: 1.0 train_size: 1000 test_size: 64 train_size: 500 test_size: 0 scripts/exps/format/svg/free_think/env_config.yaml +9 −3 Original line number Diff line number Diff line Loading @@ -3,9 +3,12 @@ env1: env_config: prompt_format: free_think split: train dino_weight: 0.0 dino_weight: 0.0001 structural_weight: 0.0 dreamsim_weight: 5.0 device: dino: 0 dreamsim: 0 train_size: 500 test_size: 0 Loading @@ -14,8 +17,11 @@ env2: env_config: prompt_format: free_think split: test dino_weight: 0.0 dino_weight: 0.0001 structural_weight: 0.0 dreamsim_weight: 5.0 device: dino: 0 dreamsim: 0 train_size: 0 test_size: 128 No newline at end of file test_size: 32 No newline at end of file scripts/exps/prompting/alfworld/model_config.yaml +5 −5 Original line number Diff line number Diff line Loading @@ -17,11 +17,11 @@ models: # max_retries: 3 # timeout: 60 # claude_3_sonnet: # provider: claude # model_name: claude-3-7-sonnet-20250219 # max_tokens: 150 # temperature: 0.7 claude_3_sonnet: provider: claude model_name: claude-3-7-sonnet-20250219 max_tokens: 150 temperature: 0.7 # Gemini_2.5_flask: # provider: gemini Loading scripts/exps/prompting/frozenlake/grounding/env_config.yamldeleted 100644 → 0 +0 −8 Original line number Diff line number Diff line env1: env_name: frozenlake env_config: render_mode: vision prompt_format: grounding use_accuracy_reward: false train_size: 10000 test_size: 128 No newline at end of file Loading
scripts/benchmark/service_benchmark/navigation/run.sh +1 −1 Original line number Diff line number Diff line Loading @@ -25,6 +25,6 @@ mkdir -p benchmark_results # Run benchmark echo "Running service benchmark" python -m vagen.utils_benchmark.service_benchmark --config "$SCRIPT_DIR/benchmark_config.yaml" python -m vagen.env.verify_service --config "$SCRIPT_DIR/benchmark_config.yaml" echo "Benchmark complete. Results saved to benchmark_results directory." No newline at end of file
scripts/benchmark/service_benchmark/svg/dataset_config.yaml +2 −2 Original line number Diff line number Diff line Loading @@ -5,5 +5,5 @@ svg-vision: dino_weight: 1.0 structural_weight: 1.0 dreamsim_weight: 1.0 train_size: 1000 test_size: 64 train_size: 500 test_size: 0
scripts/exps/format/svg/free_think/env_config.yaml +9 −3 Original line number Diff line number Diff line Loading @@ -3,9 +3,12 @@ env1: env_config: prompt_format: free_think split: train dino_weight: 0.0 dino_weight: 0.0001 structural_weight: 0.0 dreamsim_weight: 5.0 device: dino: 0 dreamsim: 0 train_size: 500 test_size: 0 Loading @@ -14,8 +17,11 @@ env2: env_config: prompt_format: free_think split: test dino_weight: 0.0 dino_weight: 0.0001 structural_weight: 0.0 dreamsim_weight: 5.0 device: dino: 0 dreamsim: 0 train_size: 0 test_size: 128 No newline at end of file test_size: 32 No newline at end of file
scripts/exps/prompting/alfworld/model_config.yaml +5 −5 Original line number Diff line number Diff line Loading @@ -17,11 +17,11 @@ models: # max_retries: 3 # timeout: 60 # claude_3_sonnet: # provider: claude # model_name: claude-3-7-sonnet-20250219 # max_tokens: 150 # temperature: 0.7 claude_3_sonnet: provider: claude model_name: claude-3-7-sonnet-20250219 max_tokens: 150 temperature: 0.7 # Gemini_2.5_flask: # provider: gemini Loading
scripts/exps/prompting/frozenlake/grounding/env_config.yamldeleted 100644 → 0 +0 −8 Original line number Diff line number Diff line env1: env_name: frozenlake env_config: render_mode: vision prompt_format: grounding use_accuracy_reward: false train_size: 10000 test_size: 128 No newline at end of file
